National recovery plan for the Christmas Island Pipistrelle Pipistrellus murrayi
Prepared by Martin Schulz and Linda F. Lumsden
Department of the Environment and Heritage, 2004
ISBN 0 642 55012 3
Part C: Known and potential threats
It is currently not known what has caused the recent decline of Christmas Island Pipistrelle. However, factors that may be influential and hence warrant investigation are listed below. It is likely that the decline of the species is the result of a combination of factors, and as the information available on species' biology and conservation ecology is limited, it is possible that other hitherto unidentified threatening processes may be of additional importance.
The Christmas Island Pipistrelle is a rainforest-dependent species that requires primary rainforest for roosting sites. The extensive clearfelling of primary rainforest for phosphate mining has reduced the roosting habitat available for the species compared to that available at the time of settlement. While opening up parts of the rainforest may have increased the area available as foraging habitat, roosting habitat is generally more restricted and limiting than foraging habitat for insectivorous bats. Hence it is expected that a population decline was experienced by the species in the years of intensive clearing for phosphate mining.
Habitat loss is not considered to be the cause of the recent decline in distribution and abundance as clearing of primary rainforest had not occurred during most of the time between the study of Tidemann (1985) and Lumsden et al. (1999) (clearing of primary rainforest ceased in 1987). However any additional loss of habitat may compound the other factors that are impacting on the species and are likely to be more influential now that the species has been reduced to lower population size and is in decline.
The proposed removal of phosphate stockpiles within the Christmas Island National Park may adversely affect foraging and commuting habitat, and possibly roosting habitat. Proposals currently under consideration to clear primary rainforest on vacant crown land may provide additional pressure on remaining Christmas Island Pipistrelle populations and/or reduce suitable habitat available for the long-term recovery of the species. These proposals include phosphate mining at sites in the eastern section of the island, and activities associated with developments such as the siting of a mobile phone tower on Limestone Hill, South Point; the Christmas Island airport upgrade; and Linkwater Road re-alignment north of the Christmas Island Resort area. Small numbers of pipistrelles have been sighted recently at Limestone Hill and in the proposed southern extension of the airport (Bamford & Bamford 2002; M. Bamford, pers. comm.).
- Yellow Crazy Ant Anoplolepis gracilipes
- Prey availability
- Habitat alteration
- Climatic conditions
- Vehicle-related mortality
- Decreasing population size
The Yellow Crazy Ant is a tramp species that has been recognised as among the top 100 of the "world's worst" invaders by the IUCN and the Global Invasive Species Database (O'Dowd 2002). It is currently under nomination for listing as a key threatening process under the EPBC Act and has been recognised as a key threat to biodiversity on Christmas Island (Commonwealth of Australia 2002). Due to its generalised biology it can be readily translocated in packaging material, timber, plants and soil, and continues to spread around the globe (O'Dowd 2002). It was accidentally introduced to the island some time between 1915 and 1934 (O'Dowd et al. 1999). These ants form multi-queened supercolonies, and dramatic increases in supercolony formation began in the mid to late 1990s at several widespread locations. The effect of the supercolonies is that the Yellow Crazy Ant may become the numerically dominant consumer on both the forest floor and in the canopy (O'Dowd et al. 1999, O'Dowd et al. 2003). Supercolonies range in size from several hectares to several hundred hectares, and at the height of their infestation occupied 25% of the total rainforest area on Christmas Island.
It is currently not known what impact the Yellow Crazy Ant has on the Christmas Island Pipistrelle. However, evidence indicates that the continuing spread of the ant would have deleterious consequences for the long-term viability of the species. The Yellow Crazy Ant has been recorded preying on mammals elsewhere, such as newborn pigs, dogs, cats, rabbits and rats (e.g. Lewis et al. 1976, Haines et al. 1994). The Christmas Island Pipistrelle is known to be attacked and killed by the ant: one individual captured in a harp trap set on the Martin Point Track died as a result of Yellow Crazy Ant attack in 1998 (Lumsden et al. 1999). Bats contacted by Yellow Crazy Ants that are not killed directly are likely to suffer reduced fitness due to exposure from sprayed formic acid leading to blindness and physiological stress (O'Dowd et al. 1999).
The 1998 study of the Christmas Island Pipistrelle by Lumsden et al. (1999) was undertaken at the time that supercolonies were beginning to form on the island and the potential problem they may cause was just starting to be recognised. Some of the locations that indicated a reduction in Christmas Island Pipistrelle abundance between 1994 and 1998 were in areas where supercolonies had started to form (e.g. Field 22S and The Dales; Lumsden & Cherry 1997, Lumsden et al. 1999). However, in the 1998 study, foraging and commuting activity was recorded in some areas that supported Yellow Crazy Ant supercolonies (e.g. Martin Point Track where the supercolony formed in 1997). Many sites that revealed a reduction in pipistrelle activity levels between 1994 and 1998 were in areas unaffected by Yellow Crazy Ant supercolonies. Therefore it cannot be concluded that the observed activity reductions in some sites were as a result of the Yellow Crazy Ant. However, since that time 14 of the 15 sites with high or moderate levels of pipistrelle activity were invaded by, or were within the bat's foraging range of, ant supercolonies (Appendix One).
All the roosts located in 1998 were in areas that were devoid of supercolonies at the time. The majority of roosts were situated under exfoliating bark, strangler figs or in hollows on the main trunks of rainforest canopy trees (e.g. Figures 7-10). These roost locations are directly in the path of foraging columns of ants travelling from nests on the ground to the canopy where they forage (O'Dowd et al. 1999). Consequently, such roost sites are likely to be readily accessed and investigated by Yellow Crazy Ants. In supercolonies the density of ants has been recorded at about five ants per 100 cm2 of tree bole at breast height (D. O'Dowd, pers. comm.). Some roost sites may also be potentially usurped by ants nesting in canopy or mid-strata tree hollows. It is likely that in areas infested by Yellow Crazy Ant, the Christmas Island Pipistrelle would be forced to select alternative roosts, where available. Such roosts may not provide appropriate structural characteristics to afford shelter from adverse weather conditions or predators, or provide the appropriate thermal microclimate. No maternity sites have been located but these roosts may be situated in the hollows of rainforest canopy trees. Similarly, these sites would also be susceptible to infestation and potential predation. Given the small size of the pipistrelle (adults weigh 3-4.5 g, with new-born young likely to weigh approximately 1 g) maternity sites located within supercolony areas, and in particular the non-volant young, must be considered at risk.
The Yellow Crazy Ant has a generalist diet foraging on seeds, a variety of leaf litter and arboreal invertebrates, crustacea including land crabs, reptiles, birds and mammals, throughout the day and night (e.g. O'Dowd et al. 1999, O'Dowd 2002). The maintenance of Yellow Crazy Ant supercolonies results in intense localised predation pressures resulting in the alteration of invertebrate diversity and abundance throughout all strata of the rainforest. For example, in affected areas compared to forests with no Yellow Crazy Ant supercolonies, Davis et al. (2002) recorded lower densities of litter-dwelling invertebrates, and G. Richards (pers. comm.) observed markedly less stridulating katydids in the rainforest canopy. Such predation pressures on invertebrates are likely to influence prey availability of the Christmas Island Pipistrelle, and require investigation. Reduction in flying insect numbers may result in reduced breeding success and a reduction in bat population size.
Yellow Crazy Ant infestations also lead to tree dieback with some rainforest canopy tree species more affected than others (e.g. Inocarpus fagifer), thereby potentially resulting in the alteration of the species composition of the primary rainforest (O'Dowd et al. 2001, 2003). Additionally, alteration in various invertebrate and seed-dispersing terrestrial bird populations may influence floristic composition of affected rainforest areas (Davis et al. 2002). Changes in rainforest species composition may have long-term effects on the Christmas Island Pipistrelle in terms of the availability of suitable roosts and maternity sites. For example, some canopy trees are more prone to hollow and exfoliating bark formation than others.
The Yellow Crazy Ant has the potential to alter the whole ecology of the island due to its generalised foraging and nesting habits in both disturbed and undisturbed habitats. For example, an estimated 15 to 20 million Red Crabs Gecarcoidea natalis have been killed since 1989, resulting in dramatically altered plant community dynamics (O'Dowd & Lake 1989, 1990, 1991; Green et al. 1997). The Red Crab has been described as a keystone species influencing the ecology of the rainforest on Christmas Island at a landscape level (Lake & O'Dowd 1991). The removal of the Red Crab results in increased seedling production, increased forest leaf litter accumulation, and an increase in understorey growth thereby altering the structure of the rainforest. Alteration in forest structure may significantly influence within-rainforest foraging habitat. The increased density of the understorey layer may exclude access to potential roosts in the lower forest strata. Evidence indicates that the Christmas Island Pipistrelle forages in canopy breaks including those along ecotones and caused by treefalls, and less frequently within the canopy in open situations with little mid-storey (Lumsden et al. 1999). A denser mid-canopy structure would restrict within-canopy foraging.
A priority conservation management objective of the Crazy Ant Steering Committee and Parks Australia North was to control Yellow Crazy Ant supercolonies using ant baits. An aerial baiting program has resulted in the destruction of supercolonies at all sites baited (Green 2002; Kemp 2003; M. Jeffery, Parks Australia North, pers. comm.). Follow-up ground surveys and baiting are at present being conducted in areas that previously did not support supercolonies and were therefore not targeted by aerial baiting. The impact of anticide baiting on the Christmas Island Pipistrelle, both directly through contact with the bait and indirectly through flow-on impacts on prey species, is unknown (e.g. Green 2002).
Except for one death due to Yellow Crazy Ants (see above), no instances of predation have been recorded. However the likelihood of observing the predation of a small, cryptic nocturnal bat is extremely low, and predation is probably occurring but going unrecorded. The Christmas Island Pipistrelle may be exposed to predation pressures from three categories of predators:
a) Common Wolf Snake Lycodon aulicus capucinus: This species is a recent coloniser from Southeast Asia that was first recorded in the Settlement area in 1987 (Smith 1988). It forages predominantly on lizards and occasionally small mammals on the ground or in the lower forest strata (Deoras 1978, Daniel 1989, Murthy 1990). The Common Wolf Snake is usually associated with human habitation and on Christmas Island is established around the Settlement area (Rumpff 1992, Fritts 1993) (Figure 12). Until 1998, the only record elsewhere on the island was of a population around the buildings at Grants Well in the centre of the island. In 1998 the location of a number of individuals further west indicated a range expansion for this species: a population was found at the Christmas Island Research and Education Station, 1.3 km south of Grants Well, and one snake was found active at night on the EW Baseline at the junction of the eastern arm of the Circuit Track (Cogger & Sadlier 1999, Lumsden et al. 1999) (Figure 12). This westward range extension appears to be continuing with a specimen located on 20 May 2003 at Field 22S, another 2.3 km further south-west (M. Jeffery, pers. comm.) (Figure 12). Although the Common Wolf Snake has been recorded on the edge of primary rainforest in the central parts of the island it is not known whether it is confined to the edges or is spread throughout rainforest tracts. The Common Wolf Snake is capable of climbing trees (Auffenberg 1980) and may predate on roosting bats, particularly those sheltering under exfoliating bark and Strangler Figs on the lower trunks of rainforest trees. Non-volant young left in maternity roosts at night while the adults are away foraging may be particularly exposed to predation by this snake.
Lumsden et al. (1999) considered this snake to be a likely factor in the observed decline and westward range contraction of the Christmas Island Pipistrelle. In 1984 when Tidemann (1985) recorded the pipistrelle to be widespread and common, including in the Settlement area, the snake was not yet introduced to the island. However, by the early 1990s, extremely high densities (up to 500 individuals per ha) were recorded (Rumpff 1992). In 1994 no pipistrelles were observed in the Settlement, although low levels of activity were recorded at a single site nearby (see Figure 5 ). No pipistrelles were recorded anywhere in the far north-eastern section of the island in 1998, and anecdotal evidence suggested they disappeared from the Drumsite area of the Settlement several years before (Lumsden et al. 1999). The expansion of the Common Wolf Snake into the central region of the island may account for the decline of the bat in that region (for example the marked decline in bat activity around the Christmas Island Research and Education Station) and the general westward contraction in distribution. No bats were recorded at the detector sites immediately adjacent to the three Common Wolf Snake locations in the centre of the island (Figure 12).
Introduced snakes have had devastating impacts on island fauna elsewhere (e.g. Savidge 1987, Fritts & Rodda 1998, Loope et al. 2001). For example, the Brown Tree Snake Boiga irregularis has caused the extinction of 75% of the native forest bird species and half the native lizards on Guam within 40 years of introduction (Loope et al. 2001), and reduced the Mariana Fruit Bat Pteropus mariannus population to only 100 adults, with no recruitment for a decade (Fritts & Rodda 1998). Of all the introduced predators on Christmas Island, the Common Wolf Snake is the only species for which the timing of the introduction was immediately prior to the decline of the pipistrelle and whose distribution mirrors that of the pipistrelle (Figure 12). Having evolved in the absence of arboreal predators, the Christmas Island Pipistrelle is likely to be naive to the risk of climbing snakes and would not have developed strategies to avoid such predation. The Common Wolf Snake has had serious detrimental impacts when introduced to other islands. For example, on Reunion Island it has been attributed with causing a decline in endemic mice and the near extinction of a species of gecko (Cheke 1987). Weighing less than 5 g the Christmas Island Pipistrelle is smaller than some of the other vertebrate species that the Common Wolf Snake has been recorded preying upon.
Figure 12. The distribution of the Common Wolf Snake up until 1998, in comparison with activity levels of the Christmas Island Pipistrelle at 84 detector sites sampled in 1998 (from Lumsden et al. 1999). A record of the Wolf Snake from 2003 indicates a continued westward expansion (M. Jeffery, pers. comm.).
b) Feral Cat Felis catus: This introduced predator became established soon after settlement, and is now common and widespread on the island. It is considered to pose a severe threat to native animals on Christmas Island (Commonwealth of Australia 2002). Although dietary studies have not revealed the Christmas Island Pipistrelle as a prey species (Tidemann et al. 1994), it is possible that occasional individuals are captured given their low roosting and foraging habits.
c) Black Rat Rattus rattus: This exotic species has been attributed with the extinction of bats on islands elsewhere in the world (e.g. Daniel & Williams 1984), and is thought to be a severe threat to native animals on Christmas Island (Commonwealth of Australia 2002). The Black Rat was introduced when the island was first settled, and is now common and widespread throughout the island, and occurs both in areas occupied and not occupied by the pipistrelle. However, it is possible that the Black Rat may be a contributing factor in the decline of the pipistrelle. Potential changes in the distribution and abundance of this opportunistic rat, in response to altered food resources as a result of the impacts of the Yellow Crazy Ant supercolonies on rainforest structure, may need to be considered.
Nankeen Kestrel Falco cenchroides: On mainland Australia, the Nankeen Kestrel preys primarily on terrestrial vertebrates, with bats occasionally recorded as a dietary item (Lewis 1987, Marchant & Higgins 1993). This raptor expanded its range and significantly increased in abundance on the island in the 1980s (H. Rumpff, cited in Lumsden et al. 1999). Although a bird of grasslands and other open habitats on mainland Australia, on Christmas Island this species is also widespread in areas of secondary rainforest regrowth. It is absent from extensive tracts of primary rainforest (such as in the west of the island). However, it is present along the edges and tracks through some areas of primary rainforest, using these openings as foraging locations. In 1984, Tidemann (1985) recorded the pipistrelle hawking insects along roads and ecotones during the late afternoon, several hours before dusk. Foraging by bats during daylight hours on islands elsewhere in the world has been attributed to a lack of avian predators (Speakman 1995). No daytime foraging of the Christmas Island Pipistrelle was observed in 1994 by Lumsden & Cherry (1997) or 1998 by Lumsden et al. (1999) suggesting a temporal shift in foraging behaviour. Such a change in behaviour may be the result of predation pressure, with emergence shifting to dusk when predation risk is lower.
The Nankeen Kestrel was widespread across the island (in 1998), both in areas that the pipistrelle had disappeared from and in areas where it was still relatively common (Lumsden et al. 1999). Therefore, it is unlikely that predation by this species is the primary cause of the decline, however, the possibility that it is a compounding factor cannot be ruled out, and requires investigation.
The Christmas Island Pipistrelle has not been recorded as a prey item of the Christmas Island Goshawk Accipiter fasciatus natalis or Christmas Island Hawk-Owl Ninox natalis (Hill 2002a, b; Hill & Lill 1998). However, it is possible that both species may opportunistically prey on this bat. The relationship between the pipistrelle and these potential natural predators is not likely to have altered recently, and hence they are not considered to be the cause of the recent decline.
Unknown factors may be altering the densities of prey available to the Christmas Island Pipistrelle. Preliminary dietary studies have indicated a range of flying invertebrates are taken as prey items. Further investigations, however, are required to determine if the species is an opportunistic feeder or shows dietary specialisation, and if this varies throughout the year. As discussed above, Yellow Crazy Ant supercolonies appear to have resulted in the reduction of invertebrate diversity and abundance. Alteration to flying insect numbers may result in reduced breeding success of the pipistrelle, leading to a reduction in population size.
In the 1960s drill lines were bulldozed across the island at 120 m intervals for phosphate mining exploration. This resulted in the clearing of 354 separate lines with a total length of 506 km (Lumsden et al. 1999). The Christmas Island Pipistrelle is an edge specialist targeting forest ecotones and gaps within the rainforest canopy. In 1984 Tidemann (1985) commonly observed bats flying along open drill lines. By the mid-1990s, the combination of storm damage and the regeneration of vegetation along many of the drill lines resulted in the loss of this temporary foraging niche. The loss of this habitat may have caused a local reduction in population numbers. It does not, however, account for the apparent abundance of this bat at first settlement (e.g. Andrews 1900) or the westward contraction in range of the pipistrelle.
Cyclones have been documented to severely impact bats on islands (e.g. Craig et al. 1994, Gannon & Willig 1994, Rodriquez-Duran & Vazquez 2001). A severe storm in March 1988 damaged significant areas of primary rainforest. The impact of this natural event on the roosting, maternity and foraging areas of the pipistrelle is unknown.
The effects of drought as experienced in 1997 and early 1998 on the Christmas Island Pipistrelle are unknown. It is likely that such conditions restrict prey numbers and may influence the thermal properties of roosts resulting in a population decline. Although forest fires are uncommon on the island, during recent extended dry periods in 1994 and 1997, fires occurred in terrace rainforest. The effects of forest fire on the Christmas Island Pipistrelle is unknown, but may result in direct adverse impacts due to the loss of roost sites (particularly exfoliating bark on tree trunks), and indirectly by affecting invertebrate populations. There may be an increased potential for wildfires in primary rainforest as a result of falling debris from wayward or failed launches from the proposed Asia Pacific Space Centre.
The Christmas Island Pipistrelle commonly forages along roads from close to ground level to above canopy height within and along the ecotone of primary rainforest and secondary rainforest regrowth (Tidemann 1985, Lumsden & Cherry 1997, Lumsden et al. 1999). Small rainforest bat species are known to be the victims of roadkills elsewhere (Schulz 2000). Currently the incidence of vehicle-related mortality (e.g. from night haulage trucks associated with phosphate mining) is unknown. Although not considered a major cause of mortality, increased night-time traffic levels along roadways may result in an increase in vehicle-related mortality, especially in the western section of the island, due to the construction of the Immigration, Reception and Processing Centre and associated infrastructure. If population numbers were high, deaths due to vehicles would probably be inconsequential. However, as numbers decrease, any additional deaths have a greater impact.
Although there is no evidence for disease in the Christmas Island Pipistrelle population the possibility that the decline in the species is due to an epidemic resulting from an introduced pathogen cannot be ruled out. It is believed that the extinction of the two native species of rats on Christmas Island was due to an introduced pathogen carried by the Black Rat (Pickering & Norris 1996). There was no obvious external sign of ill health in the pipistrelles caught during the 1994 or 1998 studies. If disease was a factor it could be expected that animals on the edge of the distribution, where the decline was most evident, would have been in a poorer condition (as indicated by lower body weights) than individuals in the west of the island. However, there was no difference between the two areas in the body weight of individuals captured.
Current evidence suggests that the Christmas Island Pipistrelle is declining in both distribution and numbers. A small population size increases the risk of extinction through inbreeding depression and stochastic events (Caughley & Sinclair 1994).
Potentially, all primary rainforest and secondary rainforest regrowth, which together provide key critical diurnal roost, maternity roost and foraging habitat, may be considered as areas under threat:
- Habitat Loss: Loss of any area of primary rainforest, secondary rainforest regrowth and rehabilitation sites will result in the elimination of currently utilised habitat or areas of habitat required for the long-term recovery of the species.
- Yellow Crazy Ant: Potentially all disturbed and undisturbed forest habitat may be under threat, depending on the success of control measures taken and maintained by the Crazy Ant Steering Committee and Parks Australia North. While most of the large supercolonies have now been controlled, numerous small supercolonies are developing in plateau areas (D. O'Dowd, pers. comm.), which if not controlled may impact on the pipistrelle in the future.
- Predators: The identified suite of potential predators occurs throughout the island in a range of habitats occupied by the Christmas Island Pipistrelle as foraging or roosting habitat. The extent of areas occupied by the Common Wolf Snake away from the Christmas Island Research and Education Station, Grants Well and the Settlement area are poorly known. It is not known whether this snake ranges into primary rainforest away from roads and disturbed edges.
- Prey Availability: Foraging occurs in a wide variety of habitats. However, it is not known which of these represent core foraging habitats and whether the species specialises on specific flying insects at various times of the year.
- Habitat Alteration: The regrowth of the drill lines bulldozed in the 1960s is occurring in primary rainforest and secondary rainforest regrowth throughout the island.
- Roadkills: Vehicle-related mortality is likely to occur on main thoroughfares between the Settlement and phosphate mining sites or developments in the south and west of the island. Increased mortality may occur as a result of the Immigration, Reception and Processing Centre, as sections of the road between this Centre and the Settlement pass through areas of high bat foraging activity (Figure 2).
- Climatic Conditions and Disease: All areas of known and potential habitat may be affected by these threatening processes.
- Decreasing Population Size: Since this declining species is restricted to Christmas Island all areas of habitat currently supporting pipistrelles may be affected by the risks associated with small population size.
Since the cause of the decline and westward contraction of the Christmas Island Pipistrelle is still to be established, it must considered that the entire population is under threat. Christmas Island supports the only known location of the Christmas Island Pipistrelle. Therefore conservation of this population is essential to the survival of the species.